联系我们

论文查询结果	返回搜索

论文编号：	8975
作者编号：	1120120787
上传时间：	2016/12/13 15:01:03
中文题目：	云服务组合优化与协同管理研究——云服务经纪商视角
英文题目：	Study on Cloud Service Portfolio Optimization and Coordinating Management——Perspective of Cloud Service Broker
指导老师：	焦媛媛
中文关键字：	云服务经纪商；质量评价；云服务组合；协同管理；信息熵
英文关键字：	Cloud services brokerage; Quality evaluation; Cloud service composition; Collaborative management; Information entropy
中文摘要：	云服务模式是对企业传统信息化模式的颠覆，无论是任何规模的企业或利益相关者，都可以从云端获取所需的任何软硬件资源和信息等服务资源。企业信息存储、管理、计算和应用需求完全可以在云端完成，即在云计算环境下，实现一切皆服务（Everything as a Service），或称泛在服务。而企业员工可以在任何时间、地点，使用各种类型的IT终端联接云端，在自身权限内，实现数据的访问、存取和运算等目的。这是全文研究的技术背景。 随着云计算技术、用户认识和政策法规的发展，越来越多的企业和个人用户逐渐接受更多的云计算服务。云服务商业正努力开发更多功能实现的云服务，以满足广大企业业务流程和个人生活等各领域越来越广泛多样的IT服务需求，从而在供给端有种类多样丰富的云服务池。云服务经纪商的业务主要是为广大企业和个人用户提供服务质量（Quality of Service，QoS）增强服务，而面对云服务供需两端越来越庞大的同行竞争和业务群体，使云服务经纪商也面临越来越大的竞争压力。这样的竞争压力使得经纪商无论是在原子云服务选择、云服务组合优化还是云服务水平协议全生命周期管理各方面，都会面临越来越多的困难和复杂过程，是比传统web服务环境下的服务选择和服务组合优化及管理问题更复杂的系列问题。如何科学评价云端服务的综合质量水平，优化服务组合构建效率和优化云服务水平全生命周期管理，是云服务经纪业务必须面对的重要课题。 针对上述问题，本文主要做如下四个方面的研究工作： （1）云服务经纪商视角的原子云服务质量水平评价 本文首先从云服务的安全保障水平、成本、服务性能和客户反馈信息四个方面内容评价云服务商及其开发的云服务总体质量水平，这四个方面分别反映了云服务质量水平的不同方面。需要特别指出的是：云服务开发成本也能体现出云服务商的实力，一般技术实力雄厚的开发商，其云服务开发成本反而低，并且服务供应越多，越能降低成本。而服务开发成本是影响服务价格的一个重要方面。最后基于这四个方面的指标构建了一个云服务质量水平动态模糊综合评价模型。 （2）客户感知云服务组合关键QoS指标提取 为更深入的了解终端客户对云服务技术和产业发展现状的认识及主要关注QoS维度，本文结合一些信息技术信任管理成果和技术接受模型，构建了一个云服务QoS特征对客户感知云服务满意度影响模型，并基于大量相关企业界人员的半结构化和结构化调查获得的数据，对这一模型假设进行统计论证研究，最后得出云服务的第三方审计水平、信誉度、可用性、治理水平和响应时间为客户最关注的云服务质量QoS指标。 （3）云服务组合优化计算技术研究 结合调查统计研究得出的五个关键QoS指标，本文进一步研究云服务组合优化计算问题。这一问题是云服务水平协议管理生命周期的起始协商阶段主要问题。云服务经纪商只有快速生成更具个性化的、综合水平良好的云服务组合，才能确保更好的云服务水平协议管理全过程，并更好地应对大量并发云服务组合需求。本文构建了一个帕累托改进粒子群优化算法模型应用于快速搜寻计算更优QoS的服务组合。模拟实验结果表明，该模型具有的计算速度和服务组合QoS提升效果。 （4）云服务经纪业务协同管理问题研究 云服务经纪业务管理全过程都包含大量的跨组织管理协同问题，特别是在云服务水平协议管理全生命周期中。本文首先详细论述了云经纪业务价值链各组织之间的业务协同内容，在这基础上结合价值网理论和协同管理理论，构建了一个云服务经纪业务价值网络协同管理理论概念模型，并基于一个单案例研究对该模型进行修正调整。最后基于案例研究得出的五个重要序参量和信息熵理论构建了一个云服务经纪价值网络协同度评价模型。
英文摘要：	Cloud service model is the subversion of traditional enterprise informatization patterns, business of any size or stakeholder, can obtain any required software, hardware resources and information service resources from the cloud. Almost any IT or DT (digital technology) business affairs, including information storage, management, computing, and application requirements etc, can completely be done in the cloud, which means that all Service can be realized in the cloud (Everything as a Service). And employees, in their own jurisdiction, can connecte in the cloud for date access and operation through any IT terminal at any time and place. These are the technical background of the researches of the full text. With the development of cloud computing technology, user`s knowledge, policies and regulations, more and more enterprises and individual users gradually accept more and more cloud computing services. Also, the cloud Service supplier`s business are working to develop more and more functions implementation of cloud services, to meet more and more wide variety of IT Service demand of enterprise business processes and individual life in various fields, so there forms a rich diversity cloud service pool on the supply side. The main business of cloud services brokerage is to provide Quality of Service (QoS) enhanced Service for enterprises and individuals, cloud services brokerages face much huge peer competition on the service supply side and more and more business groups on the demand side. Such competition pressure makes the brokers face more and more difficult and complicated process in the choice of atom cloud services, cloud services portfolio optimization or full lifecycle management aspects of cloud service level agreements (slas), which are series of problems of more complex then than traditional web service environment. So, how to scientifically evaluate the overall quality of the cloud service level, optimize service portfolio construction efficiency and optimize the whole life cycle of the cloud service level management, which are a series of important subjects of cloud services brokerage business must face. According to the above problems, this paper mainly researching the following four aspects: (1) Atom cloud service quality evaluation of the cloud service broker perspective This article first evaluate cloud services developer and the overall quality level of their cloud services from four aspects: the security level, cost, service performance and customer feedback, these four total different aspects respectively reflects the different quality level aspects of cloud services. In particular, cloud services development costs also can reflect the strength of a cloud service developers, generally, the stronger technical strength the developer has, the less cost the cloud services developed, and the more supplying of cloud service, the more reduction can made in the cost. As the cost of service development is an important aspect of the service price. Finally based on these four aspects of indicators this thesis built a dynamic fuzzy comprehensive evaluation model for the evaluation of cloud service quality level. (2) Key QoS indicators extracting from customer perception cloud service composition For better study the terminal customer understanding of cloud services technology and industry development present situation and main focus QoS dimensions, this dissertation combined with some of the information technology achievements of trust management and the technology acceptance model, to built a customer perception satisfaction model to measure the influence of cloud service QoS characteristics, and based on a large number of data obtained from the related business persons` semi-structured and structured investigation. Demonstration research statistical hypothesis of the model reflect that, the third party audit of cloud service, credibility, and availability, level of governance and response time are five most concern cloud service QoS index of the customer. (3) Research of Cloud service portfolio optimization computing technology Combinating the five key QoS metrics from statistical survey study, this dissertation further research problems of cloud service portfolio optimization. This is a main problem at the starting negotiation phase of the life cycle of a cloud service level agreements (slas) management main. Cloud brokers should only quickly generate more personalized, comprehensive level of the good combination of cloud services, can they ensure better cloud service level agreements (slas) in the whole process of management, and can they better deal with a large number of simultaneous cloud service portfolio requirements. This paper builds a pareto improved particle swarm optimization model for fast search calculation and better QoS service portfolio finding. Simulation experiment results show that the model has a certain promotion effect on computation speed and service portfolio QoS. (4) Cloud services brokerage business management studies The cloud services brokerage`s business management process contains a large number of cross-organizational collaborative problems, especially in the whole life cycle of cloud service level agreements (slas) management. This paper discusses in detail the collaborative content between cloud brokerage and other organizations in the business value chain, then，on the basis of the value network theory and collaborative management theory, to build a theory concept model of cloud services brokerage business value network collaborative management, and based on a single case study to adjust the model. Finally, based on five important order parameter from the case study and the information entropy theory, we build a evaluation model of cloud services brokerage value network coordination degree.
查看全文：	预览  下载(下载需要进行登录)